Statistics

Cheating With Polls Down Under. Guest Post by Stephen Dawson

wmbriggs.com.24

Today’s post is from our friend Stephen Dawson, who writes at hifi-writer.com. I scheduled this in advance, knowing I’d be suffering jet lag.

Professor Will Steffen from the Australian National University is one of the two or three most prominent climate scientists in Australia. The other day he lent his expertise to a current controversy in my home town, Canberra. We (lucky us!) are getting light rail. A survey shows the Canberra population leaning against it: 46.3% of Canberra residents are against it, versus 38.8% in support. But Dr Steffen and his colleague came to the rescue. In this oped they turned that opposition around, into 33.2% against, 51.9% in support, a clear majority in support!

He and Professor Barbara Norman (also an environmental scientist) from Canberra University managed this feat by a simple expedient: they excluded conservative voters from the sample (in Australia the Liberal Party is the relatively conservative one):

In our reanalysis, we used all the percentages reported in the Canberra Times article in terms of level of support for light rail according to intended voting patterns. We then removed the intended Liberal voters from the analysis, giving a total of 980 remaining respondents to the poll, comprising the categories Labor, Greens, Others and Undecided in terms of intended voting pattern.

Of course, they had solid statistical reasons for doing so.

The very low 15.9 per cent of intending Liberal voters who support light rail are indicative of an issue that has become excessively polarising along partisan political lines. Such a strong skew also has statistical implications for the poll itself, and can easily generate a misleading impression of what the poll numbers are actually showing. In particular, the overall result of weak support for light rail could be highly skewed by the view of intended Liberal voters, who make up slightly less than one-third of the total number of residents polled.

Got to get rid of those outliers!

We all owe these authors a debt. Normally statistical manipulations with a view to gaining a particular outcome are hidden. It’s rare for them to be laid out so obviously. Australia is in good hands as the good professor:

…is currently a Climate Commissioner with the Australian Government Climate Commission; Chair of the Antarctic Science Advisory Committee, Co-Director of the Canberra Urban and Regional Futures (CURF) initiative and Member of the ACT Climate Change Council.

Editor’s Note: Reader Michael Whelan fills us in on “climate” scientist Will Steffen, the man responsible for the statistics.

Will Steffen (born 1947) is an american chemist. He was the executive director of the Australian National University (ANU) Climate Change Institute and a member of the Australian Climate Commission until its abolishment in September 2013.[1] From 1998 to 2004, he was the executive director of the International Geosphere-Biosphere Programme, a co-ordinating body of national environmental change organisations based in Stockholm.[2]

This is a prime example of cherry picking data to confirm a prejudice, makes you wonder about validity of his climate research.

Categories: Statistics

12 replies »

  1. Get John Cook to consult with the analysis and you can get support up to 97%. He’s pretty good at surveys.

  2. Gary

    Thanks for stealing 97% of Briggs’ readers thunder (and that’s even accounting for Briggs’ “outliers”.

    Hey new pun for Statisticians vs. Farmers

    A farmer is out()standing in his field

    A statistician can out()liar us all…
    …not that good…maybe one our resident genius can do one better?

  3. He reckons it’s become politically skewed, but only on one side. His first contention could be right but with it actually skewed towards the other side. The inability to consider an alternate possibility is typical of a practitioner of the mysteries called climate science. He gets paid for this junk.

  4. Gary

    Thanks:

    The Statistician is Normal, everyone else is Skewed (screwed?)

    Is also appropriate

  5. Classic technique.

    All right thinking people support X.
    Some people do not support X.
    Those who do not support X are not right thinking people.
    We have no need to count those who do not support X because they are not right thinking people.

  6. If you think it’s okay to lie for a good of cause, it’s probably better to lie about that too and not let people know what you are doing.

  7. Max got in before me, and expressed the bias issue better than I would have done.
    American Liberals (socialists) hold the true & enlightened moral and political perspectives. By simple inference this means anybody who holds different views must be a combination of three things – of lesser intelligence, of some blinkered perspective or out right lying for their own personal perspective. Australian Liberals (conservatives) have errant political beliefs, so this demonstrates they must have impaired judgement for those three reasons.
    I live in Manchester England. In 2008 we had a referendum on extending light rail plus the bus network. The $4.5bn cost was to be funded by a congestion charge. 80% voted against. The light rail (cost $1.5bn) went ahead anyway. More expensive than the buses and takes twice as long as the train to central Manchester. It is always breaking down and there are 14 months of disruption as they spend >$200m on relocating a city center stop and intersection. But it is a socialist city, so they know best.

  8. “We then removed the intended Liberal voters from the analysis…”

    The reason why Liberal voters are in the majority against it, is most likely because the bulk of their taxes will go towards funding it. The voters on very low wages, etc., who generally aren’t taxed or pay very little tax, see light rail as more free stuff.

  9. It seems to me that there are exciting possibilities open to all kinds of technical problems in both the hard and soft sciences, if we start to utilize the mathematical techniques pioneered by climate scientists. Imagine what can be achieved if we apply their inventiveness to major construction projects, nuclear power plant management, and integrated circuit design.

  10. I live in Canberra so the new math applied to the light rail poll is close to home. But hey – there is another great example of left wing math from the USA. I have just posted
    Two examples of left-wing math
    http://www.warwickhughes.com/blog/?p=3851

    The proposed light rail only services from Civic to Gungahlin in the ACT far north – so why should the larger population southside be keen to pay for this. A good question for a poll might be along lines – do you support paying for the Civic to Gungahlin light rail through increases in your rates and rents?

Leave a Reply

Your email address will not be published. Required fields are marked *